Planning and allocation of bandwidth is known in packet-oriented networks. The term bandwidth is understood to mean an expected or actual value of a volume of data exchanged per time unit between network elements.
In packet-oriented networks, frequent use is made of bandwidth reservation between two or more network elements. Such bandwidth reservation is implemented, for example, for an enhanced priority data exchange, or for a data exchange that provides for a real-time or quasi real-time transmission. The latter transmission is employed in particular for communication using an RTP (real-time protocol), for a VoIP (voice over Internet protocol) communication connection, for example.
In the case where a logical call setup is initially directed to several alternate destination network elements, subsequent to a successful call setup with one of the destination network elements, in order to then maintain a single connection to this network element, a conventional bandwidth reservation proves to be inefficient in this respect, since a reservation is initially made for all alternate destination network elements. As a result, transmission capacities are unnecessarily committed for a considerable period of time. Similarly, a call setup from one of several alternate source network elements, from which a single call is sent out following a successful call set-up with one of the source network elements, proves to be just as inefficient.